All docs

AI agents

Use Rivelant from your AI agent (MCP)

Connect Claude Code, Codex, Cursor, ChatGPT and other agents to Rivelant's MCP server, sign in with OAuth or an API key, and fix connection problems.

Rivelant runs an MCP server, so AI agents that support the Model Context Protocol can read and work with your Rivelant data: site audits, rankings, Search Console data, keywords, backlinks, AI visibility and your Google Business Profile.

  • Server address: https://app.rivelant.com/mcp
  • Transport: streamable HTTP (a remote server, nothing to install)
  • Sign-in: OAuth in your browser, or an API key

The quick way

Open the page for your agent on Integrations and click the setup button. It copies a short prompt. Paste it into your agent, and the agent adds Rivelant, opens sign-in and checks the connection, then runs a first site audit.

The prompt points your agent at rivelant.com/install.md, the step-by-step instructions written for agents. Each agent also has its own shorter version, like /install/claude.md.

Set it up yourself

Add https://app.rivelant.com/mcp as a remote (HTTP) MCP server named rivelant, using your agent's own command or settings. The exact steps for Claude Code, Codex, Cursor, GitHub Copilot, Gemini CLI, OpenClaw, Hermes, ChatGPT, the Claude app and Grok are on each agent's Integrations page.

Chat apps like the Claude app, ChatGPT and grok.com add servers as connectors in their settings. An agent can't change those settings for you, so add the address there yourself.

Signing in

OAuth (recommended). The first time your agent connects, it opens Rivelant's sign-in page in your browser. Approve access and you're done. If you don't have an account yet, you can create one on that page and you'll come straight back.

Access lasts while your agent keeps using it: the sign-in renews itself, and it ends after 30 days without use. Then you sign in again.

API key. Use one only if your agent can't open a browser, for example on a server:

  1. In Rivelant, open the menu at the bottom of the sidebar, choose Settings (your account settings, not the project's), and create a key under API keys.
  2. Store it in your agent's secret settings or an environment variable such as RIVELANT_API_KEY. Never paste it into a chat or commit it to a repository.
  3. Have your agent send it as the header Authorization: Bearer <key>. (x-api-key: <key> also works.)

Rivelant doesn't support device-code sign-in, so agents on machines without a browser need an API key.

What the agent can do

The agent signs in as you and sees the projects you can see, nothing else. It picks Rivelant's tools by itself; you just ask in plain words, for example:

  • "Audit my website and tell me what to fix first."
  • "Which searches are we close to winning?"
  • "Which questions does ChatGPT answer without naming us?"
  • "Which tracked keywords dropped this week?"

Reading your data is free. Actions that fetch new data, like keyword research or a new rank tracker, use credits from your plan, and the agent gets a cost estimate first. Agents are asked to check with you before anything over 2,000 credits. Rivelant can't buy credits or change your plan from an agent.

If it doesn't work

ProblemWhat to do
The agent says Rivelant needs authentication, or you get a 401Run the agent's sign-in step again (for example /mcp in Claude Code). With an API key, check that the key is still in your settings and sent in the header.
The server is added but no Rivelant tools appearMost agents load new servers when a session starts. Reload them (for example /mcp), or restart the agent.
Your agent can't add remote MCP serversIt doesn't support remote MCP yet. Try one from Integrations.
Sign-in opens but failsMake sure you're signed in to the right Rivelant account in that browser, then try again.

Still stuck? Email support@rivelant.com with your agent's name and the message you see.